Network bursts and network oscillations are important in cognition, memory and in neuropsychiatric disease (including epilepsy and the ketamine model of memory loss and psychosis). Network bursts and oscillations are amenable to study in brain slices, from wild-type and from genetically modified animals, with application of appropriate electrical stimulation, various neuroactive drugs, or ionic or pH manipulations. Understanding of mechanisms requires biologically faithful detailed models, due to the complexity of the system: there are a) non-linear intrinsic membrane properties, b) chemical synaptic interactions (on many time scales, and spatially distributed over the neurons); c) gap junctions, both between dendrites as well as at newly discovered axo-axonal sites. All of the above contribute to the shaping of network behavior. Presently available models can explain the neuronal firing patterns during certain experimental epilepsies, but not the initiation of a seizure - perhaps due to omission of gap-junction-mediated effects prior to the seizure. Available models also do not explain important aspects of network oscillations in connexin-36 knockout mice, in which inter-neurons become electrically uncoupled. The work in this proposal seeks to extend existing hippocampal network models, so as to incorporate simultaneously: accurately described intrinsic properties, and also both synaptic and gap-junction mediated interactions between neurons. Such models will be studied in collaboration with electro-physiologists, who examine network bursts and oscillations induced by metabotropic glutamate and cholinergic agonists, kainate, alkalinization, and other means. In addition, new single-cell models will be developed of thalamic and neo-cortical neurons, with the aim of studying network bursts and oscillations in thalamocorticat circuits. This work has the potential of suggesting novel sites for drugs intended to suppress seizure initiation, and to anticipate side effects that drugs can exert on putatively normal network behavior. The work will be coordinated with research aimed at developing simpler (i.e. single-compartment and/or analytically tractable) models that can uncover basic mathematical principles.